AI Specialist (Copilot)
Bechtel View all jobs
- Gurgaon, Haryana
- Permanent
- Full-time
- Relocation Authorized: None
- Telework Type: Full-Time Office/Project
- Work Location: Gurugram
- Identify AI Opportunities: Work closely with different departments to identify potential AI use cases that can solve business problems and create value.
- Use Case Development: Develop detailed AI use cases, including problem statements, objectives, expected outcomes, and success metrics.
- Feasibility Analysis: Conduct feasibility studies to assess the technical and economic viability of proposed AI use cases.
- Solution Design:
- Develop and maintain data ingestion workflows from varied sources (APIs, databases, file systems).
- Design and implement Copilot Agents with autonomous planning, multi-turn reasoning, memory, and tool/function calling.
- Configure system prompts, guardrails, policies, and content filters to ensure safety and reliability.
- Build actions, plugins, and custom connectors to integrate with REST APIs, Dataverse, Dynamics 365, SharePoint, Azure SQL, and internal systems.
- Implement RAG (Retrieval-Augmented Generation) using embeddings, vector indexes, and enterprise knowledge sources.
- Optimize prompt engineering for accuracy, determinism, and cost-performance tradeoffs.
- Develop end-to-end orchestration flows in Copilot Studio, including triggers, branching, grounding, and hand-offs to tools.
- Configure data connections, knowledge sources, and security contexts via Entra ID and RBAC.
- Create reusable templates and components for rapid solutioning and scale.
- Documentation: Maintain comprehensive documentation of AI use cases, including requirements, design, implementation, and performance metrics.
- Continuous Improvement: Stay up-to-date with the latest advancements in AI and continuously seek opportunities to improve existing AI solutions and processes.
- Enforce Responsible AI policies: safety, fairness, privacy, and content filtering.
- Education: Bachelor’s degree in computer science, AI, Classic Machine Learning models, Data Science, or a related field. Master’s degree preferred.
- Experience: 4+ years of experience in AI/ML engineering, software engineering, or data engineering.
- 1–3 years building and deploying LLM-based assistants or conversational/agentic systems.
- Proven track record shipping production copilots with monitoring and lifecycle controls.
- Experience with Semantic Kernel or LangChain.
- Vector databases/search: Azure AI Search, Pinecone, Weaviate, or similar.
- MLOps/LLMOps: model/versioning, automated evaluations, CI/CD for AI assets.
- Certifications (any of):
- Microsoft Certified: Azure AI Engineer Associate
- Microsoft Power Platform Developer Associate
- Copilot Studio: agent configuration, orchestration, grounding, connectors, actions/plugins.
- Copilot Agents: planning, tool calling, memory, safety guardrails, evaluation.
- Azure OpenAI: GPT family, embeddings, function calling; prompt design & optimization.
- Integrations: REST APIs, OAuth, JSON schemas; Azure API Management is a plus.
- Azure Services: Azure Functions, Azure Storage/ADLS, Azure AI Search/Vector index, Event Grid/Service Bus (nice to have).
- Programming: Proficiency in Python and SQL; Power Automate/Power Fx exposure is a bonus.
- Hands-on expertise with Azure OpenAI Studio (prompting, fine-tuning, deployments)
- Experience with MLflow, MLOps, CI/CD pipelines.
- Solid understanding of data modeling, distributed systems, and cloud-native architectures.